Dartmouth (following MIT’s lead) is reinstating use of the SAT in admissions. In this posting --a primer on SATs and admissions – and a rejoinder to Princeton Prez's defense of diversity admissions.
After going test optional during the pandemic, MIT and Dartmouth have returned to requiring applicants to submit SATs. What are the benefits – and the potential costs?
During the pandemic, most colleges went “test optional”, but now MIT and Dartmouth have reinstated use of the SAT. Their decisions were based on evidence that requiring applicants to submit SAT scores improved the ability of the schools to admit the students who were most likely to succeed academically – and that the scores were particularly helpful when assessing minority applicants.
Why did so many colleges go test optional?
Ostensibly the colleges did not want to require applicants to have to sit in large crowded rooms taking the SAT or ACT at a time when COVID risks were high. However, it’s clear that, for a couple of reasons, many schools were pleased to be able to use COVID as an excuse for no longer requiring applicants to submit their test scores:
1. Some have claimed – without direct evidence – that the tests are biased against racial and ethnic minorities.
2. The large differences in the average test scores of African American applicants admitted to Harvard and UNC, in comparison with the scores of Asian Americans, was an important source of evidence that the two schools were discriminating against Asian Americans, leading to the Supreme Court decision overturning affirmative action in college admissions. It will be much easer for schools to defend themselves against charges of racial discrimination in admissions if they do not require SAT or ACT scores as part of the application/admission process.
Are the tests racially biased?
There is no question that there are racial differences in mean scores on the tests. Asian Americans, on average, score higher than whites, who score higher than African Americans – and the differences are not small. Some have claimed, based on that evidence alone, that the tests must be racially biased. However, such a view requires acceptable of the premise that ANY racial difference between average levels of performance in any domain must be a result of racism. This is the argument that Ibram X. Kendi (author of How to be an Antiracist) has made. However, that claim literally begs the question by presuming a cause (racism) for an outcome in the absence of direct evidence.
How could we know if the tests are racially biased?
Colleges originally chose to use the tests because student scores were predictive how how well students were likely to perform at the schools – higher scores being predictive of (on average) higher first year GPAs and a higher likelihood of the student eventually graduating. If the tests are racially biased, then it should be the case that African Americans with much lower scores than whites or Asians should perform as well as higher scoring whites or Asians in the same programs at the same universities. In other words, it should be the case that when an African American student receives a score of, for example, 1200, that score for that individual is as predictive of successful performance in college as would be the case for an Asian American high schooler who achieves an SAT score of 1350. However, the best contemporary evidence indicates that the predictive validity of the tests holds across racial groups, and there is no good evidence that underrepresented racial minorities with lower scores are likely to perform at college at the same level as whites of Asians with higher scores.
Is there necessarily a trade-off between a focus on academic merit and diversity in admissions?
In his recent article in The Atlantic, Princeton president Christopher Eisgruber has argued that elite universities can “have it all”, that is, that they can maintain or increase racial diversity without any sacrificing of excellence. Jerry Coyne has pretty well eviscerated Eisgruber’s arguments in a recent Why Evolution Is True posting, so I’ll just make a couple of points here.
First – yes – promoting diversity will not decrease academic excellence of a student body IF a school has been acting to severely limit applications and/or admissions for individuals who are members of particular ethnic, religious, or racial groups. For example, at one time, Princeton – and Harvard and Yale and other Ivy League schools – had a functional quota on how man Jews they would admit. Ending that racist system – by focusing more heavily on academic merit than on family background when making admissions decisions – increased the schools’ academic excellence AND their diversity (if you accept that admitting more Jews to a previously heavily WASPY white school involves an increase in diversity). Indeed, any efforts to expand the applicant pool has the potential to increase both diversity and excellence – as long as all students are judged against the same standards when making admissions decisions. And ironically, and as Dartmouth and MIT have discovered, requiring applicants to submit SAT scores is particularly valuable as a tool for discovering which applicants from less advantaged backgrounds are most likely to be able to be successful if they attend the university.
Secondly — However if the efforts to increase diversity involve the admission of students with much poorer academic qualifications, and if those poorer qualifications are in fact predictive of a lesser ability to perform at a high level at the university (which is the case for students with SAT scores hundreds of points below the average for students at any particular university), then in that case there WILL NECESSARILY BE a tradeoff between academic excellence and increasing the diversity of the student body. And it is difficult to understand how Eisgruber can write as if that isn’t true, because that is the entire point of affirmative action – to admit students – in the name of increasing diversity – who would not otherwise be qualified to be admitted to the university.
If colleges return to requiring SATs AND, in the interests of increasing diversity, admit students from some racial and ethnic groups who have much lower than average SAT scores (for that school), won’t that open the schools up to further lawsuits (now that affirmative action has been banned).
Yes. But apparently, Dartmouth and MIT are now willing to take that risk or are willing to admit many more Asian students and many fewer African American and Hispanic students. It will be interesting to see what transpires at these and the other selective schools over the next few years — whether they simply ignore or superficially “work around” the Supreme Court’s ruling while still including race per se as a factor in admissions (as some schools have stated they will), or whether they will accept that as long as there are racial disparities in students’ average academic qualifications in high school, there will necessarily be some groups who are “underrepresented” at our most selective colleges.